Machine-driven meta-research : the application of big data approaches to map openness and transparency across the biomedical research literature
Abstract/Contents
- Abstract
- Recent concerns about the transparency and reproducibility of science have led to several calls for more open and transparent research practice. However, with almost 25,000 new biomedical articles published per week, manually mapping and understanding changes in transparency is unrealistic. In this dissertation, we develop the computational tools to study biomedical literature at a large scale. We first extract and study all preprints ever published in the biomedical literature to understand the extent to which the community has embraced this medium of quick and direct communication of information. We then extract and study all retracted articles to understand the extent to which the scientific community and the lay public are made aware of such retractions. We finally extract the entire open biomedical literature, develop automated algorithms to map indicators of transparency within this literature and characterize indicator distribution across time, fields of science and journals. Our results indicate that researchers are rapidly embracing opportunities to share their research as preprints and that these preprints receive substantial attention. They also indicate that instead of acting as a warning against inaccurate results, retractions of popular articles do not receive much attention and often inadvertently promote the original article. Finally, our results indicate that transparency of information is increasing over time, but uptake of some indicators (e.g. data sharing) lags behind others (e.g. conflicts of interest) and that practice varies enormously between fields of science. This work has also established two open source packages and an integrated database of the entire open biomedical literature to facilitate future research in the field. Our automated approach enables large-scale analyses that would have otherwise been unrealistic and establishes the foundations for computational and big data approaches to studying scientific research itself
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2020; ©2020 |
| Publication date | 2020; 2020 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Serghiou, Stylianos |
|---|---|
| Degree supervisor | Iōannidēs, Iōannēs P. A |
| Thesis advisor | Iōannidēs, Iōannēs P. A |
| Thesis advisor | Baiocchi, Michael |
| Thesis advisor | Goodman, Steven N |
| Thesis advisor | Sainani, Kristin |
| Degree committee member | Baiocchi, Michael |
| Degree committee member | Goodman, Steven N |
| Degree committee member | Sainani, Kristin |
| Associated with | Stanford University, Program in Epidemiology |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Stylianos Serghiou |
|---|---|
| Note | Submitted to the Program in Epidemiology |
| Thesis | Thesis Ph.D. Stanford University 2020 |
| Location | electronic resource |
Access conditions
- Copyright
- © 2020 by Stylianos Serghiou
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...